What problem does it solve? Turning vague feature ideas into unambiguous, testable requirements is hard, and unstructured notes break down when handed to an AI agent or another developer. This Skill captures a single feature as a structured spec.yaml contract with stable requirement IDs, cross-cutting constraints, and dependency tracking. ## Core Features & Use Cases - Guided requirement interview: Asks targeted questions about behavior, users, functional requirements, and constraints, then groups answers into UPPER_SNAKE component keys with numbered, testable requirements. - Create and update flows: Detects existing specs by name or content similarity, then either creates a new spec.yaml from a template or updates an existing one while keeping requirement IDs (ACIDs) stable. - Dependency inference: Scans other specs in the project to infer requires relationships and manages the draft status lifecycle that hands off to the planning step. - Use Case: You want to add CSV export to your app. Run the skill, answer a short interview, and get specs/csv-export/spec.yaml with grouped requirements, constraints, and inferred dependencies ready for planning. ## Quick Start Ask the AI to create a spec for user authentication using the mini-sdd-spec skill.
